The present invention relates to vehicle location-based services (LBS), and more specifically, to vehicle domain data streaming processing systems.
The increasing development of vehicle telematics, sensor and mobility technologies have given rise to the prevalence of mobile computing systems and location-based services (LBS). In order to enhance vehicle telematics applications, it is typically necessary to stream data of a moving vehicle with real-time LBS. Examples of these enhanced vehicle telematics applications include, for example, on-line event alerts, congestion monitoring, real-time driving route guidance, driver approaching alerts for ride-sharing application, and autonomous navigation.
Conventional attempts to process data streaming for facilitating LBS applications include implementing existing internet-based frameworks such as, for example, Hadoop, Storm, Spark etc. These internet-based frameworks, however, present various unique challenges. For instance, unlike conventional internet-based data processing methods that typically receive data from a stationary or substantially stationary terminal or node, the vehicle domain provides large amounts of constantly moving terminal devices operating under various conditions. In addition traditional GPS/3G/4G signals are unstable and easily affected by many factors. For example, weather conditions, high buildings, elevated roads, tunnels, bridges etc., may contribute to signal errors, interruptions, and delays.
Internet-based frameworks are also typically designed to process independent requests. The vehicle domain, however, is different in the sense that requests are not independent. That is, a single vehicle may request a multitude of requests at a single instance such as, for example, contextual map and road network information, events, traffic/weather conditions, dynamic navigation instructions, and detour instructions. Therefore, fast and large shared memory access is necessary to process the requests in parallel so as to avoid delivering stale data to the end-user.